app.py

import gradio as gr
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
import torch
import spacy
import pytextrank
from sumy.parsers.plaintext import PlaintextParser
from sumy.nlp.tokenizers import Tokenizer
from sumy.summarizers.luhn import LuhnSummarizer
from sumy.summarizers.lex_rank import LexRankSummarizer
import nltk

nlp = spacy.load('pt_core_news_sm')
nltk.download('punkt')

#WHITESPACE_HANDLER = lambda k: re.sub('\s+', ' ', re.sub('\n+', ' ', k.strip()))

model_name="GiordanoB/mT5_multilingual_XLSum-sumarizacao-PTBR"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

def summarize_HUB(input, method, max_length, min_length, no_repeat_ngram_size, num_beams):
    if method == "Pure mT5":
        return sumarize_mT5(input, max_length, min_length, no_repeat_ngram_size, num_beams)
    if method == "Luhn":
        return sumarize_Luhn(input)
    if method == "LexRank":
        return sumarize_LexRank(input)
    return "tchau"

def sumarize_Luhn(input):
    summ = ''
    summarizer = LuhnSummarizer()
    parser = PlaintextParser.from_string(input, Tokenizer("portuguese"))
    summary_1 = summarizer(parser.document, 3)
    
    for sentence in summary_1:
        summ = summ + ' ' + str(sentence)
    summ2 = ''
    summ2 = summ.replace('\n', ' ').replace('\r', '')
    return summ2

def sumarize_LexRank(input):
    summ = ''
    summarizer = LexRankSummarizer()
    parser = PlaintextParser.from_string(input, Tokenizer("portuguese"))
    summary_1 = summarizer(parser.document, 3)
    
    for sentence in summary_1:
        summ = summ + ' ' + str(sentence)
    summ2 = ''
    summ2 = summ.replace('\n', ' ').replace('\r', '')
    return summ2

def sumarize_mT5(input, max_length, min_length, no_repeat_ngram_size, num_beams):
  for i in range(0,14):
    input_ids = tokenizer(
        input,
        return_tensors="pt",
        padding="max_length",
        truncation=True,
        max_length=512
    )["input_ids"]

    output_ids = model.generate(
        input_ids=input_ids,
        max_length=max_length,
        min_length=min_length,
        no_repeat_ngram_size=no_repeat_ngram_size,
        num_beams=num_beams
    )[0]

    response = tokenizer.decode(
        output_ids,
        skip_special_tokens=True,
        clean_up_tokenization_spaces=False
    )
  return response

gr.Interface(fn = summarize_HUB, 
    inputs=["textbox",gr.Radio(["Pure mT5","Luhn","LexRank"]), gr.Slider(50, 200, step=1, value=200),gr.Slider(25, 100, step=1, value=75), gr.Slider(1, 10, step=1, value=1), gr.Slider(1, 10, step=1, value=1)], outputs=["text"]).launch()